Tilburg University:
AI-ready?

Educate, Encourage, Experiment, Invest

3/23/23

1. Let’s talk about AI in education

Who am I?

Disclaimer

  • Not an AI researcher (but: I love technology!)
  • Researcher in business & economics (but: involved 10+ faculty and staff members, across schools and divisions)
  • Today: focus on teaching, not research (full deck: https://tiu.nu/aiready)

AI is everywhere

In daily life

  • Search and recommendation (e.g. Bol.com)
  • Face recognition (e.g., iPhone)
  • Navigation (e.g., Google Maps)
  • Voice assistants (e.g., Alexa)

At Tilburg University

  • Plagiarism detection (e.g., TurnItIn)
  • Text editing (e.g., “Editor” in Word or Grammarly)
  • Automatic messages (e.g., Teams)
  • Voice assistants (e.g., Outlook’s “play my email”)

And then came ChatGPT…

What’s new?

  • General-purpose AI (vs. specialized AI)
  • Language model (seems to understand)
  • Trained on massive data (with internet connection at Bing)
  • Conversational (iterative input, “creative” output)

How to use it?

  • Open chat.openai.com
  • Login & refresh until you’re in
  • Start a conversation!

Examplary use of ChatGPT (for preparing my exam): Prompt, ask & refine

Examplary use of ChatGPT (for preparing my exam): Prompt, ask & refine

2. How is AI currently used at Tilburg University?

How we use AI in teaching

  • Course design and management
    • Creating and revising exam questions
    • Generating and/or improving syllabi and lecture outlines
  • Content delivery
    • Finding examples and case studies
    • Creating and improving slides, generating video subtitles
  • Automation
    • Connect different systems
    • Semi-automated grading

We leverage AI to efficiently improve education through engaging, multimodal, and high-quality content.

Lecturing using AI: text-to-speech + avatar animation (by Anna Paley, TiSEM)

How our students use AI

  • Searching information (e.g., much faster than Google)
  • Note-taking during class (e.g., voice-to-text; typing & filling in gaps)
  • Drafting, fixing or annotating computer code (e.g., “I have a bug in this code - do you spot it?”)
  • Ideation and problem solving (e.g., “which R package to use for X?”)
  • A “virtual” teacher, 24/7, immediate response (e.g., “why is X important?”, “can you explain how X works?”, “I want to learn more about X. Where to start?”)
  • Editor (e.g., finding good titles & writing abstracts, feedback on writing style, developing paragraph structures, synthesizing; CV writing)

Text editor popular among students (Quillbot)

3. Novel applications of AI

Future use cases

  • Low-hanging fruits
    • Improving assessment (e.g., more formative testing, generation of questions)
    • Aligning course content with “blended” learning strategy (e.g., updating syllabus, introducing more interactive moments)

  • Challenging but possible with today’s technology
    • Chatbots (e.g., student help desk, IT help desk, but also study & career coach)
    • Automating assessment (e.g., written feedback, automated grading)
    • “Recycling” existing course content (old lectures) by creative cutting or format conversions

Risks and opportunities

Risks of embracing AI

  • ChatGPT = copying and cheating?
  • “Will students still learn X?”
  • How to adapt teaching methods and assessment?
  • Quality checks when scaling up?
  • Less physical interaction?
  • Impact on staffing?

Opportunities of embracing AI

  • Using AI is relevant for our students’ future and society at large
  • Learning how to “use” an AI is not self-evident: that’s education, too!
  • Offer novel ways to deepen knowledge and foster curiosity
  • Allow researchers to stay productive
  • Aligns with and promotes our strategy (e.g., blended learning, taking courage )

Initial investment of time, effort, and money. But can enhance quality of education and research.

4. Making Tilburg AI-ready!

Making Tilburg AI-ready!

    1. Educate staff and faculty (e.g., use of ChatGPT, use cases, no-coding and coding tools)
    1. Encourage use of AI (e.g., campus license, build community)
    1. Experiment what works (and what does not; including opening up IT infrastructure)
    1. Invest in “moonshot” projects
    • Personalization of admission requirements for courses and programs
    • “Algorithmic” and interdisciplinary study programs + blended learning
    • Real-time scheduling (including students’ study rooms)

Conclusion

  • Students and some faculty already explore the use of AI in education (and research)
  • AI offers numerous benefits, but also risks that need to be addressed
  • To fully leverage AI, we need a clear “AI strategy” (with goals and budget)


Let’s embrace AI with courage and curiosity to increase the quality of our education and research.

QR Code
Download slides at https://tiu.nu/aiready.